Content uploaded by Richard J. Jensen
Author content
All content in this area was uploaded by Richard J. Jensen on Dec 04, 2015
Content may be subject to copyright.
Isozyme Variation in Oaks of the Apostle Islands in Wisconsin: Genetic Structure and Levels
of Inbreeding in Quercus rubra and Q. ellipsoidalis (Fagaceae)
Author(s): Stan C. Hokanson, J. G. Isebrands, Richard J. Jensen, James F. Hancock
Source:
American Journal of Botany,
Vol. 80, No. 11 (Nov., 1993), pp. 1349-1357
Published by: Botanical Society of America
Stable URL: http://www.jstor.org/stable/2445720 .
Accessed: 14/02/2011 10:40
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=botsam. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Botanical Society of America is collaborating with JSTOR to digitize, preserve and extend access to American
Journal of Botany.
http://www.jstor.org
American
Journal of
Botany
80(11): 1349-1357. 1993.
ISOZYME VARIATION IN OAKS OF THE
APOSTLE ISLANDS IN WISCONSIN: GENETIC
STRUCTURE AND LEVELS OF INBREEDING IN
QUERCUS RUBRA
AND Q. ELLIPSOIDALIS
(FAGACEAE)'
STAN C. HOKANSON,2 J.
G. ISEBRANDS,3 RICHARD J. JENSEN,4
AND
JAMES F. HANCOCK2'5
2Department of
Horticulture,
Michigan
State
University,
East Lansing,
Michigan 48824;
3Forestry
Sciences
Laboratory,
Rhinelander,
Wisconsin
54501; and
4Department
of
Biology,
St. Mary's
College, Notre
Dame, Indiana 46556
Isozyme
variability was examined in populations
representing
the red
oak complex
(Quercus
subg.
Erythrobalanus)
on
an island
archipelago and adjoining
peninsula
in Lake Superior,
near
Bayfield,
Wisconsin.
A concomitant
study of mor-
phometric
variation
described
in
the
companion
manuscript,
revealed a continuum in
leaf
morphology
extending
from
an
interior
mainland site to the
outermost island. The existence of
this clinal variation presented an ideal opportunity to
examine the
genetic
structure of
a hybrid
population
along with the
putative
progenitor
species. Dormant leaf bud samples
were
collected from
specimens of
Quercus
rubra L., Q. ellipsoidalis Hill,
and their
putative
hybrids from
three
islands
and
two locations
on the
peninsula.
Acorns
were collected
from
some of these same
trees from one
peninsula location
and two
islands. Twelve putative
enzyme
loci from six
enzyme
systems were
analyzed. Allele
frequency
data indicated little
differ-
entiation between
populations. Mean FST values for the
adult trees and acorns were
0.042 and 0.020. Genetic
identities
according to Nei ranged from
0.958 to 0.999. Despite these
high
levels of
genetic
similarity,
the
populations
appeared to
be highly inbred as indicated
by
positive
mean FIT
values of
0.183 and 0.373 for
the adult
trees
and acorns.
Estimates of
migration rate
per
generation
(Nm) for the
adult
trees was 5.70, a value that
is low
when
compared to estimates for
other
plant
species
with
similar
life
history
characteristics.
Considerable effort
has
been directed toward establish-
ing the relationship between life history traits and the
electrophoretic diversity within various plant species
(Brown, 1979; Hamrick, Linhart, and Mitton, 1979; Gott-
lieb, 1981; Loveless and Hamrick, 1984; Hamrick and
Godt, 1989). Comparisons
have been
made between
plant
species with similar ecological and life history attributes
to determine
whether
such
species
maintain similar
levels
of genetic variability and/or genetic structure. Hamrick
and Godt (1989) used the following eight ecological and
life history traits
in
making such comparisons: taxonomic
status, regional distribution, geographic range,
life
form,
mode of reproduction, breeding system, seed dispersal
mechanism, and successional status.
This
and earlier stud-
ies (Hamrick, Linhart, and Mitton, 1979; Loveless and
Hamrick, 1984; Nevo, Beiles, and Shlomo, 1984) uncov-
ered significant correlations between ecological
and life
history traits and patterns of genetic diversity and struc-
ture.
These
reviews concur that
species such
as
oaks,
which
are
long-lived, outcrossed, wind-pollinated, and charac-
teristic
of
the
later
stages
of
succession, should
maintain
I
Received for publication
17
December 1991; revision accepted
4
June
1993.
The authors thank Dr. Robert Brander, Julie Van Stappen, and
their staffat the
Apostle
Islands
National Lakeshore for
logistical support
and
access
to the
islands; Gary
Garten
and Dave
Buckley
for
help
with
collections;
Mike
Kwantes and
Paco
Moore for
expert computer
assis-
tance;
Dr.
Steve
Krebs for
insightful
discussion
and
instruction
on
iso-
zyme analysis; Karen Ludlam
for
generous
technical
assistance;
and
Drs.
Steve
Tonsor
and
Tim
Holtsford
for
thoughtful
comments
on early
drafts of
this
manuscript.
5Author
for
correspondence.
higher
levels
of
genetic
variation
within
their
populations
than between
them.
While
the
genetic
structure
of
herbaceous
plant
species
has been
widely
studied
(Hamrick
and
Godt, 1989),
only
limited
information
is available on woody angiosperm
species (Bousquet, Cheliak,
and LaLonde, 1987, 1988;
Surles, Hamrick, and Bongarten,
1989; Schnabel and
Hamrick,
1990; Schwarzmann and Gerhold,
1991). This
is unfortunate,
because such information is crucial
for
developing
programs to
protect and preserve
the
genetic
resources
of
plant
species. Presently,
several
species of
the
northern
hardwood
forests,
including
Quercus
rubra,
are
being
subjected to
extreme
logging
pressure (Laursen
and DeBoe, 1991). Additionally,
it is now thought that
certain environmental
perturbations,
such as global
warming and acid rain,
could threaten
the
existence of
some tree
species (Ledig, 1992). To make responsible
decisions
concerning the
possible impacts of
such
pres-
sures,
we
must build a more
substantive
base of
knowledge
regarding the
population
genetic
structure of
woody
an-
giosperm
species.
One of
the most
determinant
factors
concerning
the
population
genetic
structure
of
an outcrossing
wind-pol-
linated
species such as oaks are the
levels of
gene
flow.
While much research
has been directed
toward
under-
standing
the
dynamics
of
gene
flow
(Levin and Kerster,
1974; Moore, 1976; Slatkin,
1981, 1985; Handel, 1983;
Ellstrand,
1988),
the
actual
amount
of
gene flow
via
pollen
or
seed
is
one of the lesser
known
parameters
concerning
natural
plant
populations
(Levin, 1984;
Hamrick,
1987).
Due to a general lack of
suitable
markers,
direct
mea-
surement of gene flow is often difficult
(Ellstrand
and
1349
1350 AMERICAN JOURNAL
OF BoTANY [Vol.
80
Marshall, 1985). It is generally
agreed
that
the
most
ef-
ficient
method
for
determining
actual
gene flow
is
through
the
use of
isozyme
marker
genes
detected
with
electro-
phoretic
techniques.
The distribution
of oak populations on the Apostle
Islands
in Lake
Superior
presented
us with the
opportunity
to study
the
population
genetics
of members
of
the
red
oak complex.
The islands
and
adjacent
Bayfield
Peninsula
(BP) (Jensen
et al., 1993) support
natural
populations
of
only
two species of
subgenus
Erythrobalanus.
Morpho-
logical
characters
of leaves
and
acorns
indicate
that
Quer-
cus
rubra
L. (northern
red
oak) predominates
on
the
out-
ermost
island
of
the
archipelago
(Outer
Island),
while Q.
ellipsoidalis
Hill (northern
pin
oak) predominates
in the
peninsula
interior.
Between
these
distal
populations
are
populations
that
can be characterized
as being
interme-
diate or
hybrid
in
nature
(Jensen
et al., 1993).
This pattern
of morphological
variability
gave us a system
in which
to pursue
the degree
of
interaction
between
Q. rubra
L.
and Q. ellipsoidalis.
We
were
interested
specifically
in the
following
questions:
1)
How is genetic
variability
distrib-
uted
within
red
oaks in a zone of
overlap;
and 2) How
much and how far
do genes move among these
island
populations
of oak?
MATERIALS AND METHODS
Collections.
-Five populations
of oaks were
studied
in
the
Bayfield,
Apostle
Islands National Lakeshore region
in northeastern
Wisconsin
(Jensen
et
al., 1993). Samples
were
collected
on five separate
trips
to the
region.
Two
trips
were made
in
June
of
1989,
and
one each
in
October
1989, March 1990, and May 1990. BP collections
were
made along the roads that
defined
the east and north
perimeters
of the Chequamegon National Forest.
Pen-
insula
perimeter
(PP) collections
were
made on
the
public
roads
that
skirted
the
perimeter
of the
BP. At both
sites,
we sampled
the
first
mature
oak of any
taxon
sighted
at
0.4-km
intervals
along
the
roads.
Collections
on Oak Is-
land (OK) were
made on
a north
to
south
trail
that bisected
the island.
The Stockton
Island (STK) and Outer
Island
(01) collections
were done on trails
that ran
from
south-
west
to
northeast
the
length
of each
of the
islands.
These
populations
were
sampled by walking
vigorously
for
5
minutes
and collecting
the
first mature
oak of
any
taxon
sighted
within
20 yards
of
the
trail.
Samples collected
in June 1989 consisted
of
rapidly
expanding
leaves 15-75 mm
in
length.
On the
subsequent
trips,
three
to
five
pencil-sized
branches
containing
dor-
mant
buds
and acorns
were removed
from individual
trees
using
a
shotgun.
Buds
were left
on small
twigs
while
acorns
and leaves
were
removed.
Samples
were
placed
in
labeled
zip-lock
bags,
stored
within 3 hours
on ice in a cooler,
and on return
to East Lansing,
Michigan,
placed in a
refrigerator
at 2.5 C until
enzymes
were extracted.
Mean
sample
sizes
per
locus
for trees
and
acorns
at
each location
are listed
in Table 2.
Enzyme
extraction
- Samples
were
macerated
with
pes-
tles
in
chilled mortars using
the
Soltis
phosphate
grinding
buffer
(Soltis
et
al., 1983). Prior
to
grinding,
one spatula
tip of insoluble
polyvinylpolypyrrolidone (PVPP, Sigma
#P-6755)
completely hydrated
with
the grinding
buffer
was added. Approximately 10 mg
of
dormant
bud was
ground per
specimen.
For
acorns, 35
to 40
mg
of
cotyledon
tissue, removed
from the
cap
end ofthe
acorn,
was
ground
per sample.
Leaves rapidly
lost
their
enzyme activity
and
had to be prepared within days
for
electrophoresis,
while
dormant buds
could
be stored for
up
to
2 months
before
extraction
with no appreciable
loss in activity.
Acorns
ground 11 months after harvest also had high enzyme
activity.
After
grinding, the resulting slurry was absorbed through
nylon mesh
onto
4 x 1
1-mm
wicks cut from #3
Whatman
chromatography paper.
Wicks sufficient
for four electro-
phoretic
examinations were placed in Corning
96 well
disposable ELISA plates that were double wrapped
in
cellophane,
bagged
in
zip-lock bags,
and stored
at -80 C
until
electrophoresed.
Only enough
wicks
for
one
analysis
were placed
in
a plate, so that the wicks were
never
taken
out of the freezer
until
they
were to be analyzed.
Well-
resolved
isozymes
were obtained from
samples
stored
up
to 14 months under these conditions.
Electrophoresis
- Experiments
with
a number
of
gel/
electrode buffer
systems
revealed that the
pH 8.3 lithium
borate, tris-citrate
system (Scandalios, 1969), and the
morpholine-citrate
pH 6.1 system (Clayton
and Tretiak,
1972) clearly
resolved the largest number of enzymes.
Ten-millimeter-thick 6.1 gels were typically run for 7
hours at 55-65 milliamps, and approximately 250 volts.
Six-millimeter-thick 8.3 gels
were run for 6 hours at 50
milliamps
or
until
300 volts
was
reached. Slices from the
8.3 Scandalios system
were
stained
for
phosphoglucose
isomerase (PGI) (Vallejos, 1983) and leucine amino pep-
tidase
(LAP) (Soltis
et
al., 1983).
Slices from the
6.1 mor-
pholine-citrate
system were stained for 6-phosphogluco-
nate dehydrogenase
(6-PGDH) (Conkle et al., 1982),
isocitrate
dehydrogenase (IDH) (Soltis et al., 1983), mal-
ate
dehydrogenase
(MDH) (Vallejos, 1983),
and
shikimate
dehydrogenase
(SKDH) (Soltis
et
al., 1983). All staining
assays
were
conducted
as cited
except
for LAP. The sub-
strate used was L-leucine-B-napthylamide
HCL (Sigma
#L0376) which was dissolved directly
in the
buffer so-
lution. Once the bands stained
clearly,
the slices were
rinsed
in
1
%
acetic acid solution
and fixed
in
a 50%
ethanol
solution. Slices were then
bagged
in zip-lock bags and
refrigerated
at
4 C for later
analysis.
Bands
were read
in
the conventional manner
with those
loci migrating farthest
from
the
origin being designated
as number
one,
the next farthest number
two,
etc. Within
a locus the fastest allele was named one, the
next two,
etc.
(Figs. 1, 2). Because we made no controlled
crosses
to analyze
segregation
of
the
isozyme banding patterns,
all allele
and loci designations
are
putative.
Our initial
experiments
indicated that resolution
be-
came increasingly poor as leaves expanded,
while dor-
mant buds
gave
consistently good
results.
Thus,
we
report
results from
only dormant bud or acorn samples. The
acorns
analyzed
in this
study
were
collected from
adult
trees
that were also analyzed electrophoretically.
Five
acorns were
electrophoresed
from
each of
these
parent
trees.
November 1993] HOKANSON
ET AL.
-ISOZYME VARIATION
IN OAKS OF THE
APOSTLE
ISLANDS 1351
* y%* ..* -A * -
21-
PGI A B
21
22\
MDH A B
1300_
SKDH A B C
Fig.
1. Electrophoretic patterns
for three of the six
enzyme systems
used
in
this
study.
Numbers
to the
left of the
photographs
indicate
putative
loci.
Superscript
numbers indicate
putative
alleles of each locus.
PGI was scored as a dimeric
enzyme
encoded
by
two
loci. Locus one
was considered
monomorphic. Lane
A
was
heterozygous
for alleles three
and
five of
locus two. Lane
B
was
heterozygous
for alleles one and
five
of locus two.
MDH was scored as a dimeric
enzyme
encoded
by
three
loci.
In
this
photograph
all
individuals were
homozygous
for
allele one
of locus one. Lane
A
was
heterozygous
for alleles one and two of locus
two. Allele one of locus three has
co-migrated
with allele two of
locus
two.
Lane B was
homozygous
for allele two of locus two. Allele
one of
locus three has
co-migrated
with allele two of locus two. SKDH was
scored as a monomeric
single
locus
enzyme.
Lane
A
was
homozygous
for allele two.
The
second band
was
thought
to be a plastid
form of the
enzyme
and not
scored. Lane
B shows a heterozygote
for alleles
three
and five. The faint bottommost band was
considered a plastid
form of
allele five and not scored. Lane C was
homozygous
for allele five.
Statistical
analysis
-Allele frequencies, average het-
erozygosities (direct
and
estimated), percent loci poly-
morphic,
mean
alleles per
locus, F-statistics, and genetic
identity according to Nei (
1978) were calculated using the
BIOSYS-
1
program,
release 1.7, which is adapted for the
personal computer (Swofford
and Selander, 1989).
11
i3>
14
15\
21
IDH B A DC E
2212
231%
%
31K
32-.
LAP ABC D __
Fig.
2. Electrophoretic
patterns
for the
remaining
three
of six
enzyme
systems
used in
this study.
Labeling
remains
the
same as on
the
previous
page.
IDH was scored
as a
dimerc enzyme
encoded by
three
loci.
Lane
A was
heterozygous
for alleles
one and
three of
locus
one. Lane
B
was
heterozygous
for
alleles four
and
five of locus
one. Lane
C was
hetero-
zygous
for alleles
one
and two of
locus
two. Lane
D was
homozygous
for allele
one
of locus
three.
Lane
E was
heterozygous
for alleles
two
and
three of
locus
three. 6PGDH was
scored
as a dimerc
enzyme
encoded
by
two loci.
Lane A was
homozygous
for
allele one
of
locus
one.
Lane B was
heterozygous
for
alleles
one and
two of locus
one.
Lane
C was heterozygous
for
alleles
one
and
three of
locus
one. Lane
A
was
homozygous
for allele
two
of locus
two.
Lane
D was
heterozygous
for
alleles
one
and
two
of locus
two.
LAP was
scored
as
a
monomerc,
single
locus
enzyme.
Only
the
uppermost
locus
was scored
for this
study.
Lane
A was homozygous
for
allele
two.
Lane
B
was
homozygous
for
allele
three.
Lane
C
was
homozygous
for allele
one.
Lane D was
heterozygous
for alleles
two and
four.
Direct
or
observed
average
heterozygosities
were
cal-
culated
by
adding
the
number
of
heterozygous
individuals
at each
locus
in
the
population,
dividing
this
number
by
the
total
number of
individuals
in the
population,
and
averaging
this value
over
loci.
Calculations
of
average-
estimated
heterozygosity
were
based on
Hardy-Weinberg
1352 AMERICAN JOURNAL OF BOTANY [Vol. 80
TABLE 1. Allele
frequencies at 12 putative
enzyme loci for five
adult and three acorn
populations of Quercus
studied on the Apostle
Islands and
vicinity.
BP PP OK STK 01
Locus Allele Adlt Acor Adlt Acor Adlt Acor Adlt Acor Adlt Acor
PGI- 1 1 1.00 - 1.00 - 1.00 - 1.00 - 1.00 -
PGI-2 1 0.000 - 0.014 0.000 0.000 0.000 0.000 - 0.000 0.000
2 0.030 - 0.000 0.011 0.000 0.027 0.000 - 0.000 0.007
3 0.790 - 0.688 0.819 0.583 0.818 0.656 - 0.667 0.691
4 0.010 - 0.036 0.011 0.046 0.009 0.021 - 0.028 0.026
5 0.170 - 0.261 0.160 0.370 0.145 0.323 - 0.306 0.276
LAP-1 1 0.340 - 0.176 0.457 0.186 0.474 0.427 - 0.225 0.244
2 0.300 - 0.485 0.283 0.539 0.386 0.302 - 0.250 0.411
3 0.180 - 0.250 0.207 0.186 0.096 0.177 - 0.425 0.256
4 0.180 - 0.088 0.054 0.088 0.044 0.094 - 0.100 0.089
6PGD1 1 0.930 - 0.623 0.693 0.641 0.696 0.670 - 0.706 0.710
2 0.040 - 0.098 0.034 0.141 0.107 0.085 - 0.147 0.117
3 0.030 - 0.279 0.273 0.217 0.196 0.245 - 0.147 0.173
6PGD2 1 0.000 - 0.007 0.022 0.010 0.009 0.010 - 0.000 0.000
2 1.00 - 0.986 0.978 0.990 0.982 0.990 - 0.975 0.994
3 0.000 - 0.007 0.000 0.000 0.009 0.000 - 0.025 0.006
IDH-1 1 0.448 - 0.341 0.389 0.255 0.219 0.372 - 0.306 0.253
2 0.000 - 0.000 0.000 0.000 0.000 0.032 - 0.000 0.019
3 0.354 - 0.333 0.211 0.431 0.430 0.351 - 0.361 0.474
4 0.000 - 0.024 0.000 0.010 0.000 0.000 - 0.000 0.000
5 0.198 - 0.302 0.400 0.304 0.351 0.245 - 0.333 0.253
IDH-2 1 0.000 - 0.000 0.000 0.020 0.000 0.064 - 0.050 0.058
2 0.000 - 0.101 0.133 0.059 0.175 0.021 - 0.000 0.168
3 1.00 - 0.899 0.867 0.922 0.825 0.915 - 0.950 0.744
IDH-3 1 0.020 - 0.087 0.167 0.069 0.069 0.022 - 0.000 0.094
2 0.950 - 0.884 0.813 0.931 0.879 0.957 - 1.00 0.906
3 0.030 - 0.029 0.021 0.000 0.052 0.022 - 0.000 0.000
MDH-1 1 0.980 - 0.980 0.989 0.980 0.991 0.990 - 1.00 0.980
2 0.020 - 0.020 0.011 0.020 0.009 0.010 - 0.000 0.020
MDH-2 1 0.000 - 0.014 0.011 0.000 0.000 0.000 - 0.289 0.006
2 1.00 - 0.986 0.989 1.00 1.00 1.00 - 0.711 0.994
MDH-3 1 0.950 - 0.958 0.989 0.980 0.983 1.00 - 1.00 0.977
2 0.050 - 0.042 0.011 0.020 0.017 0.000 - 0.000 0.023
SKDH1 1 0.048 - 0.030 0.000 0.010 0.000 0.000 - 0.000 0.000
2 0.269 - 0.523 0.716 0.539 0.552 0.510 - 0.556 0.636
3 0.029 - 0.152 0.136 0.088 0.208 0.198 - 0.167 0.130
4 0.029 - 0.053 0.034 0.029 0.000 0.031 - 0.056 0.026
5 0.606 - 0.235 0.114 0.314 0.240 0.260 - 0.222 0.208
6 0.019 - 0.008 0.000 0.020 0.000 0.000 - 0.000 0.000
expectations.
For each locus,
allele frequencies
were
in-
serted
into
a Hardy-Weinberg
equation derived
for
the
number
of
alleles at that
locus. These values were then
averaged
over the
number
of loci within
the
population.
The percentage
of polymorphic
loci
was calculated
us-
ing a 95% criterion,
i.e., a locus was considered
poly-
morphic
only
if
the most
common allele occurred
at a
frequency
of 0.95 or less
in
the
population.
The number
of loci in
a population
that fit this criterion
was divided
by the
total
number
of
loci
in
the
population
to generate
this
percentage.
Mean alleles per
locus
were calculated
by
summing
all the
alleles across loci in a population
and
dividing
by
the total number
of loci.
F-statistics
(fixation
indices;
Wright,
1951, 1965)
were
calculated according
to Nei (1977). This procedure
mea-
sures the deviation
of
genotype
frequencies
from Hardy-
Weinberg
expected
frequencies
in a subdivided
popula-
tion.
Deviation in
heterozygosity
from the
level
expected
under Hardy-Weinberg
equilibrium
is partitioned
into
three components,
FlS, FST, and FIT. FlS describes
the
inbreeding
in individuals relative
to the subpopulations
to which they belong.
FlS
= (hs
- h.)/h,,
where ho is
equal
to the frequency
of
heterozygous
individuals
in
an island
population and h. is equal to the expected frequency
of
heterozygous
individuals
in
an equivalent
random
mating
island population.
FST is the proportion
of the deviation
from equilibrium contained
within subpopulations
and
represents a measure of the differentiation
between
pop-
ulations.
FST = (hT -hS)/hT, where hT = the expected
frequency
of heterozygous individuals
in an equivalent
random mating
total population.
FIT is a measure
of the
reduction
in heterozygosity of an individual
in relation
to the
whole
population.
FIT
can be viewed
as the
total
heterozygote deviation
from Hardy-Weinberg
equilibri-
um. It is
comprised
of both the deviation due to
nonran-
dom mating
within
island populations
(Fls), and to the
November 1993] HOKANSON
ET AL. -ISOZYME VARIATION IN OAKS OF THE APOSTLE ISLANDS 1353
TABLE 2. Mean sample
size per
locus,
percent
polymorphic
loci,
mean and effective number
of
alleles
per
locus,
and mean
heterozygosity,
direct
and expected,
for
five
adult and three acorn
populations
of
Quercus
in the
Apostle
Islands and vicinity.a
% Mean# Mean heterozygosity
Mean sample Polymorphic alleles/ Effective
# __
Population size/locus loci locus alleles/locus Direct Expected
Bayfield 49.9 (0.3) 63.6 2.7 (0.5) 1.60 0.198 0.242
(0.069) (0.085)
Peninsula 66.3 (1.8) 63.6 3.2 (0.4) 1.79 0.232 0.342
Perimeter (0.073) (0.083)
Perimeter- 43.1 (1.9) 63.6 2.8 (0.3) 1.61 0.187 0.291
Acorns (0.061) (0.075)
Oak Island 50.5 (0.6) 63.6 2.9 (0.4) 1.71 0.249 0.303
(0.084) (0.085)
Oak-Acorns 53.2 (2.6) 63.6 2.7 (0.3) 1.64 0.181 0.296
(0.058) (0.077)
Stockton Island 47.9 (0.4) 54.5 2.7 (0.3) 1.74 0.245 0.296
(0.074) (0.090)
Outer Island 19.0 (0.3) 63.6 2.4 (0.3) 1.76 0.327 0.322
(0.090) (0.089)
Outer-Acorns 79.5 (3.2) 63.6 2.9 (0.3) 1.72 0.203 0.313
(0.065) (0.080)
Mean 51.2 (16.4) 62.5 (3.0) 2.8 (0.21) 1.70
(0.07) 0.228 0.301
(0.045) (0.027)
a Numbers
in parentheses indicate standard
error.
heterozygote deviation due to the
subdivision
of the
pop-
ulation (FST); FIT = (hT- hO)/hT.
Effective number ofalleles per locus (Aep) was calculated
according to Weir (1989), where
Aep
= 1/(I -Hep). Hep,
the genetic
diversity per
locus is equal to 1 - 2;p12.
Here
pi
equals the
frequency
of
the
iPh allele in each
population.
This analysis was done to "weight" the alleles present
in
the
population.
Because the
calculation
is based on the
frequency of the
allele
in
the population
rather than
mere
presence,
more
"weight"
is
given
to alleles with a higher
frequency.
The number of migrants exchanged per generation (Nm)
was also estimated
where N
equals
the effective
population
size and m equals the
proportion
of
migrants exchanged
between populations per generation.
Nm was calculated
using
the
FST value described
previously. According
to
Wright (1931), FST = 1/(1 + 4Nm).
RESULTS
Twelve putative enzyme loci were consistently score-
able and subsequently employed
in
this
study (Fig. 1,
2).
Eleven of the 12 loci were polymorphic; however,
no
alleles were
found
to
be unique
to
any
of
the
populations
of adult trees
or acorns
except
for
the
rare Allele 1 of
PGI-2 (Table 1).
All populations, both adult
and acorn,
had 63.6% polymorphic loci except
for
the
STK adults
which were polymorphic for 54.5% of their loci (Table
2).
Mean number of alleles per locus ranged from 2.4 of
01 adults
to
3.2 for the
PP adults
(Table 2). The effective
number of alleles per locus ranged from 1.60 in the
BP
adult
population
to 1.79 in
the
PP adult trees
(Table 2).
Mean heterozygosity
in each of the populations was
lower than expected according to Hardy-Weinberg pre-
dictions, except among the
01 adult trees (Table 2). This
overall deficiency
in
heterozygotes was reflected by pos-
itive mean FIT in both adults and acorns, although there
was considerable variation across loci. The majority of
the deviation
from
equilibrium
resided
among
individ-
uals within
populations (Fls). Mean FIT in adult trees
equaled 0.183, while the mean
FIS
component was 0.147
(Table 3). Similarly, the acorn mean FIT value was 0.37 3,
while the mean
FIS
component totaled 0.360 (Table 4).
Mean migrant per generation estimates (Nm) for
adults
and acorns were substantially different.
The mean Nm
estimate for adult
populations
was 5.70 compared
to a
value of 12.25 for the acorns (Table 4). Values for the
individual loci in adults ranged from 0.84 at the
MDH2
locus to 49.75 at the
MDH 1 locus (Table 3). The acorns
showed more variation-values for individual loci
ranged
from
7.10 at the
IDH1 locus to
249.75 at
MDH3 (Table
4).
TABLE 3. Deviations from
Hardy-Weinberg equilibrium among in-
dividuals
(FS), among populations
(FST), and total deviation
(FIT),
with
migrants per generation (Nm), according
to
Wright (1931), for
11
loci from
Quercus
adults.
#
of
Locus alleles FIS FIT FST Nm
PGI-2 5 -0.023 -0.003 0.020 12.25
6-PGDHI 3 -0.111 -0.057 0.049 4.85
6-PGDH2 3 -0.017 -0.009 0.008 31.0
LAP1 4 0.452 0.447 0.046 5.18
IDH 1 5 0.070 0.081 0.012 20.58
IDH2 3 0.542 0.555 0.029 8.37
IDH3 3 0.651 0.659 0.025 9.75
MDH1 2 -0.019 -0.014 0.005 49.75
MDH2 2 -0.383 -0.065 0.230 0.84
MDH3 2 0.088 0.106 0.020 12.25
SKDH1 6 0.130 0.179 0.056 4.21
Mean 3.45 0.147 0.183 0.042 5.70
1354 AMERICAN JOURNAL OF BOTANY [Vol.
80
TABLE 4. Deviations from
Hardy-Weinberg
equilibrium
among in-
dividuals
(F1s),
among
populations
(FST), and total
deviation (FIT),
with
migrants per
generation
(Nm),
according to Wright
(1931), for
11
loci from
Quercus acorns.
#
of
Locus alleles FIS FIT FST Nm
PGI-2 4 -0.028 -0.007 0.020 12.25
6-PGDH1 3 0.087 0.093 0.007 35.46
6-PGDH2 3 -0.017 -0.012 0.005 49.75
LAPI 4 0.523 0.536 0.028 8.68
IDH1 4 0.105 0.135 0.034 7.10
IDH2 3 0.870 0.871 0.008 31.0
IDH3 3 0.793 0.796 0.016 15.38
MDH1 2 -0.016 -0.014 0.002 124.75
MDH2 2 -0.009 -0.006 0.004 62.25
MDH3 2 -0.019 -0.018 0.001 249.75
SKDH1 4 0.555 0.562 0.016 15.38
Mean 3.09 0.360 0.373 0.020 12.25
Values for
Nei's unbiased
genetic
identity revealed the
adult
and acorn
populations
to
be closely
related
in
each
pairwise
comparison
(Table 5). Identities
ranged
from
0.958 for the
comparison
of
the
BP adult
population
to
the
PP acorn
population,
to 0.999 for the
comparison
of
the
PP adult
population
to
OK adults
(Table 5). Pairwise
comparisons
of the
BP adult
population
to the other four
populations of adults and three
populations of acorns
showed the Bayfield
adults to be the most genetically
disparate
ofthe
populations
studied. The
BP
adult
identity
values,
which
ranged
from 0.958 to
0.979, contained six
of
the lowest
seven
pairwise
genetic
identity
values
(Table
5).
DISCUSSION
Identification
of species
and their
hybrids-Our survey
of 21 enzyme systems
on seven buffer
systems yielded
no
species-specific
alleles except
for
the rare PGI-21. This
was not
surprising
considering
results from
similar stud-
ies. Manos and
Fairbrothers
(1987) found that
among
the
systems
they
examined,
aside from
identifying
Quercus
palustris, isozymes
did not
separate
species
within
Ery-
throbalanus. Guttman and Weigt (1989) attributed the
lack of allozyme divergence among the red oaks to ex-
tensive hybridization between the species. Because neither
of these
earlier
studies examined populations of
Quercus
ellipsoidalis,
we had hoped to find an allele unique to that
species; this was not the case.
Whittemore and Schaal (1991) used variations
in
chlo-
roplast DNA and nuclear ribosomal DNA to study in-
terspecific gene
flow
in
five
species
of white
oak
native to
the
eastern
United States.
Their
study
revealed a species-
specific length
variant
of nuclear
ribosomal
DNA that
distinctly separated
three
species
of
white oak. Such var-
iants could prove to be useful in making an unambiguous
identification of red oak species and their hybrids.
Genetic
variability
in
Apostle region oak populations-
We found the levels of
allozyme
variation
in
the
Apostle
region oaks to be high when compared to other vascular
plant species. Hamrick and Godt (1989) found that species
have,
on average, 50.4% of their loci polymorphic, 1.96
alleles
per locus,
and 1.21 effective alleles
per
locus.
Among
the
Apostle region
oaks we found mean
values of 62.5%
of the loci polymorphic,
2.8 alleles per locus, and 1.7
effective alleles per
locus. Other
oak studies have found
alleles per locus to vary
from 1.4 to 2.5, and percent
polymorphism
from
30 to 65 (Manos and Fairbrothers,
1987; Guttman and Weigt, 1989; Schnabel and Hamrick,
1990; Schwarzmann and Gerhold, 199 1). The life history
characteristics of oaks would lead one to predict high
levels of allozyme diversity.
In particular, Hamrick and
Godt found that
geographic
distribution and breeding
system
correlate
strongly
with
variability, particularly
at
the
species
level. Thus, geographically widespread,
out-
crossing, wind-pollinated species
such as
oak
are
expected
to have
higher
levels
ofgenetic
variation than
plant species
as a whole.
The oak populations
of
Manos and Fairbrothers ( 1987)
contained the least
diversity
of
any
oak study.
Guttman
and Weigt (1989) attributed the
low
value of Manos and
Fairbrothers to the
limited
range
of
the species over which
they sampled. The implication
here
is that
by limiting
the
range
over which
the
species
is sampled,
the
degree
TABLE 5. Unbiased genetic
identity
values
according
to
Nei (1978), for
Quercus
adults and acorns
from
the
Apostle
Islands and vicinity.
Population 1 2 3 4 5 6 7 8
1 Bayfield
Peninsula
Adults
2 Peninsula
Perimeter 0.972
Adults
3 Peninsula
Perimeter 0.958 0.988
Acorns
4 Oak Island 0.971 0.999 0.976
Adults
5 Oak Island 0.971 0.990 0.993 0.986
Acorns
6 Stockton
Island 0.979 0.995 0.990 0.993 0.993
Adults
7 Outer
Island 0.967 0.987 0.976 0.984 0.976 0.988
Adults
8 Outer Island 0.967 0.996 0.987 0.996 0.993 0.993 0.986
Acorns
November 1993] HOKANSON
ET AL.
-ISOZYME VARIATION
IN OAKS OF THE APOSTLE
ISLANDS 1355
of genetic variability
is commensurately
lowered. How-
ever,
we
collected
individuals from a much more restrict-
ed range than any
of the
other studies
and found levels
of variability that
were comparable to
or higher than those
previously reported.
This indicates that
the amount of
variation present
in red oak species may
relate more close-
ly to how many individuals
are sampled
within a popu-
lation, rather than
the breadth of the
geographical area
considered.
The levels of
heterozygosity we observed
in the oaks
and acorns
in the
Apostles region
were
generally
lower
than the expected
Hardy-Weinberg
equilibrium
values
(Table 2). In only
one instance did the
observed
mean
value exceed the
expected; among the
OI adult trees
there
was a small
excess
of heterozygotes.
This observation
was
mirrored by
the positive mean FIT values for both adult
and acorn populations,
although
the individual values
varied greatly from
locus to locus (Tables 3, 4). It is not
known why the
loci were so variable,
unless they
were
differentially
affected
by
selection.
Guttman and Weigt (1989) also described
a general
deficiency
of heterozygotes among
adult trees.
Using
their
data
we
calculated
the mean levels
of
expected
and
direct
heterozygosity
from the ten species
of red oak they-
stud-
ied. These values
equaled 0.178 and 0.103, respectively.
Although
our
expected (0.301) and direct
(0.228) values
were higher, the difference
between
our two values (0.073)
was very similar to
the difference
between the values from
Guttman and
Weigt's
data
(0.07 5).
Schnabel and
Hamrick
(1990) and Schwarzmann
and Gerhold (1991) also re-
ported
a deficiency
in heterozygotes,
but the differences
between
expected
and direct
values were much
smaller
(<0.005).
Levels
ofgeneflow
among
the
Apostle
region
oaks -The
low levels of heterozygosity
found
in
the oak populations
might
be the result
of
limited gene
flow and high
levels
of
inbreeding.
Our mean estimates
of
migrants per gen-
eration
(Nm) for the
Apostle region
oaks were
high
for
plant species as a whole. However, the
level we estimated
from
the
adults (5.70) is on the low end
of the spectrum
for
species that are widely
distributed, long-lived,
out-
crossing,
and wind-pollinated.
Hamrick
(1987) presented
migration
rates
from
plant species
fitting
into
various life
history categories
based on breeding
system.
The wind-
pollinated,
outcrossed
category
was represented by
four
conifers with
Nm estimates
ranging
from 5.3 to 37.8.
Schuster, Alles, and Mitton
(1989) reported
Nm from
another
conifer,
Pinusflexilis
James (limber
pine),
as 11.1
migrants per generation.
The low Nm values of
oaks may be related to their
mode of
seed dispersal.
Acorns are
dispersed by gravity
and
animals,
while conifer seeds are
winged
and
disperse
by
the wind.
According
to
Hamrick
(1987), species
with
winged seed dispersal
mechanisms have an average
mi-
gration rate that
is at least 20 times greater
than gravity-
and animal-dispersed
species.
Additional evidence
for
low levels
of
dispersal among
the
Apostle region
oaks can be
seen at the
empirical level,
as no consistent
patterns
were observed across islands
between
adult and acorn allele frequencies
(Table 1). If
there
had been
high
levels of
pollen
flow,
allele
frequencies
in
acorn
populations
would be expected
to deviate
from
their parental
frequencies
in the direction
of adjacent
is-
land populations.
However,
in most
cases acorn frequen-
cies were
most closely
aligned with their
own adult
pop-
ulation. In fact, the acorns were almost twice as
homozygous
as the adult
populations,
as revealed
by the
FIT values (Tables 3, 4). This FIT value for the
acorns
(0.373) indicates that substantial
inbreeding
is occurring
within these
populations.
While this
heterozygote
deficiency
was quite congruous
with low
migration rates,
other elements
ofthe data
seemed
to
imply
the opposite.
Measures of unbiased
genetic
iden-
tities revealed
the oak populations in
the Apostles
region
to be
quite
similar, a pattern
often indicative
of high
levels
of
migration
(Table 5). High levels of
gene flow were
also
indicated
by the low FST
values calculated
for these
pop-
ulations
(Tables 3, 4). Among both the
acorns and
adults,
the levels
of genetic variation
between
populations
(FST)
were the
smallest component
of the overall
deviation
from
Hardy-Weinberg
equilibrium.
A similar
pattern
was ob-
served
by
Manos and Fairbrothers (1987), Schnabel
and
Hamrick
(1990), and Schwarzmann
and Gerhold
(1991).
Another
line of evidence indicative
of high levels of
gene
flow among the
Apostle region
oaks involved
pat-
terns
of
morphological
variation (see companion
manu-
script-Jensen
et
al., 1993). Using
principle component
analysis
techniques,
we found evidence
of a morpholog-
ical cline
that paralleled
the geographic
layout of
the re-
gion.
For
principle component
one, the
continuum
of leaf
morphologies
extended
from typical
Quercus ellipsoidalis
on the
left
to typical
Q. rubra on the
right. Trees from
the
BP population
clustered to the left,
01, STK, and PP
trees clustered to the right
of
center,
while OK trees
were
found
near
center. Despite the fact
that the positioning
of the
populations
on the
cline
was not always
perfect,
the
morphological
evidence
indicated
that substantial
gene
flow had occurred
from the
mainland
into the islands.
Therefore, the isozyme
and
morphological
data
indicate
two
seemingly
contradictory
trends.
Low levels of
allo-
zyme
heterozygosity
indicated
that
these
populations
are
currently
experiencing
little
gene
flow,
while
high
genetic
identities
and
morphological
patterns
suggested
that con-
siderable
hybridization
and gene flow has occurred,
at
least
in
a historical
context. One possible
explanation
for
the
lower than
expected
levels of
heterozygosity
among
our
oak populations
concerns
the
logging history
of
the
Apostles
region
tree communities.
Significant episodes
of
logging
have
occurred
on these
islands
and
the
mainland.
The last
major logging
activity
in
the islands ended
in the
1930s while
logging
continues
to this
day
in
the
Chequa-
megon
National Forest on the
BP. In all
the
populations
we sampled,
there
were
a few
remnant
stands of
virgin
trees.
It is
likely
that
some
of the
secondary growth
stands
we sampled
were
the
products
of
"bottlenecked"
regen-
eration
from
the
patchy
remnant stands
of
unlogged
trees.
If
this
was the
case, these
secondary
stands were
regen-
erated
from
a restricted
parent pool, and thus
are more
similar
in genetic
constitution
(more
homozygous)
than
those
stands found
on less
disturbed
sites. Such "bottle-
neck" regeneration
may
have contributed
to the
increased
level of
homozygosity
we noted
in the
Apostles
region
oaks.
It may also be only
in exceptional
years
that
the
BP
trees,
which
hold the most
diverse
set
of
genes
for
the
1356 AMERICAN JOURNAL OF BOTANY [Vol. 80
region,
contribute
to the pollen
pool, and thus
serve
to
increase
levels
of heterozygosity.
In two
seasons of
col-
lecting
in
the
Apostles
region,
no acorns
were found
among
the
BP adult trees.
These BP trees
exist in a sand
barrens
area. The soil is sandy,
xeric,
and quite disturbed
by
logging
activity.
The trees
are reduced
in
size
and appear
quite
stressed.
The lack
of acorns
noted
among
these
trees
may be paralleled
by
a lack of
pollen
production.
That
is, in response
to marginal
survival conditions,
the
BP
trees may
be diverting
energy
from sexual
reproduction
to mere survival or asexual reproduction.
Only in an
exceptional
year,
i.e., one with above average rainfall,
mild
winter,
minimal
browse
damage,
etc.,
will these
trees
flower and shed
pollen,
thus adding
to the
genetic
vari-
ability
noted
in
the region.
In conclusion,
we suspect
that
these
populations
are
currently experiencing
high
levels of inbreeding
due to
"bottlenecked"
regeneration,
or diminished
pollen
input
from the
region's
most genetically
diverse
population.
In
a historic
or more
episodic
context,
the populations
may
have experienced
higher
levels of pollen-mediated
gene
flow which
would account
for the high
genetic
identities.
From a management
perspective,
the
lower than
expected
levels of
gene
flow suggest
that
proximally
located
pop-
ulations of oaks must be maintained. Such a strategy
would be
critical
for the
preservation
of the species
genetic
resources
by
allowing
for periodic
mixing
of
adaptively
significant
alleles.
LITERATURE CITED
BousQuET, J., W. M. CHELIAK, AND M. LALONDE. 1987. Genetic dif-
ferentiation among 22 mature populations
of
green
alder Alnus
crispa
in central Quebec. Canadian Journal of Forestry
Research
17: 219-227.
5 , AND . 1988. Allozyme
variation
within and
among
mature
populations
of
speckled
alder Alnus
rugosa
and re-
lationships
with
green
alder
A. crispa.
American Journal of Botany
75: 1678-1686.
BROWN, A. H. D. 1979. Enzyme polymorphism
in
plant populations.
Theoretical
Population
Biology 15: 1-42.
CLAYTON, J.
W., AND D. N. TRETIAK. 1972.. Amine-citrate
buffers for
pH control
in starch gel electrophoresis.
Journal of the
Fisheries
Research
Board of
Canada 29: 1169-1172.
CONKLE, M., P. HoDGsIKss, L. NUNNALLY, AND S. HUNTER. 1982.
Starch
gel electrophoresis
of
conifer seeds: a laboratory
manual.
General
technical
report
PSW-64 U.S.D.A. Forest Service
Pacific
SW Forest
and Range Experiment
Station, Berkeley,
CA.
ELLSTRAND, N. C. 1988. Pollen as
a
vehicle
for the
escape
of
engineered
genes?
In J. Hodgson and A. M. Sugden
[eds.],
Planned release of
genetically
engineered
organisms,
s30-s32. Trends
in biotechnol-
ogy/Trends
in
ecology
and evolution
special publication.
Elsevier,
Cambridge.
5 AND D. L. MARSHALL. 1985. Interpopulation
gene flow by
pollen
in
wild
radish,
Raphanus sativus.
American Naturalist 126:
606-616.
GOTTLIEB, L. D. 1981. Electrophoretic
evidence and
plant populations.
In L. Reinhold, J. B. Harborne,
and T. Swain [eds.],
Progress
in
phytochemistry,
vol. 7, 1-46. Pergamon,
Oxford.
GuTTMAN, S. I.,
AND L. A. WEIGT. 1989. Electrophoretic
evidence
of
relationships
among Quercus
(oaks) of
eastern North
America. Ca-
nadian Journal of
Botany
67: 339-35 1.
HAMRICK, J.
L. 1987. Gene flow and distribution of
genetic
variation
in
plant
populations.
Differentiation patterns
in
higher
plants.
Ac-
ademic Press,
New York,
NY.
,
AND M. J. GODT. 1989. Allozyme
diversity
in
plant
species.
In A. H. D. Brown,
M. T. Clegg, A. L. Kahler,
and B. S. Weir [eds.],
Plant
pop'n. genetics,
breeding and genetic
resources,
43-63. Vi-
nauer,
Vunderland,
MA.
,Y. B. LiNHART, ANDJ. B. MITTON. 1979. Relationships
between
life
history
characteristics
and
electrophoretically
detectable
genetic
variation in
plants.
Annual
Review
of
Ecology and Systematics 10:
173-200.
HANDEL, S. N. 1983. Pollination
ecology, plant
population
structure,
and gene flow. In L. Real [ed.],
Pollination
biology,
163-211. Ac-
ademic Press,
Orlando, FL.
JENSEN,
R. J., S. C. HOKANSON, J. G. ISEBRANDS,
AND J.
F. HANCOCK.
1993. Morphometric
variation in oaks of
the
Apostle
Islands in
Wisconsin:
evidence
of
hybridization
between
Quercus
rubra L.
and
Quercus
ellipsoidales
Hill. American
Journal
of
Botany 80: 1358-
1366.
LAURSEN,
S. B.,
AND J.
F. DEBOE. 1991. The
oak resource
in
the
upper
midwest-implications
for
management.
Minnesota
Extension
Ser-
vice publication, #NR-BU-5663-S. University
of
Minnesota,
St.
Paul, MN.
LEDIG, F. T. 1992. Genecology:
the fitness of the
organism
and the
fitness of the environment. In S. J.
Colombo, G. Hogan, and V.
Wearn
[eds.],
Proceedings
of the Twelfth North American Forest
Biology
Workshop,
27-47. Ontario
Ministry
of
Natural
Resources,
Ontario Forest
Research
Institute,
and Forestry
Canada, Ontario
Region.
LEVIN, D. A. 1984. Immigration
in plants:
an exercise
in the sub-
junctive.
In R. Dirzo and
J.
Sarukhan
[eds.],
Perspectives
on
plant
population
ecology,
242-260. Sinauer,
Sunderland,
MA.
, AND H. W. KERSTER. 1974. Gene flow in seed plants. Evo-
lutionary
Biology 7: 139-220.
LOVELESS, M. D., AND J.
L. HAMRICK. 1984. Ecological determinants
of
genetic structure
in
plant
populations. Annual
Review
of
Ecology
and Systematics 15: 65-95.
MANOS, P. S., AND D. E. FAIRBROTHERS. 1987. Allozyme
variation
in
populations
of
six northeastern
American red
oaks
(Fagaceae: Quer-
cus
subg.
Erythrobalanus).
Systematic
Botany
12: 365-373.
MOORE, P. D. 1976. How far
does pollen
travel?
Nature
260: 388-
389.
NEI, M. 1977. F-statistics and
analysis
of
gene
diversity
in
subdivided
populations.
Annals
of
Human Genetics 41: 225-233.
1978. The theory
of
genetic
distance and
evolution of
human
races.
Japanese
Journal
of
Human Genetics 23: 341-369.
Nevo, E., A. Beiles, and R. B. Shlomo. 1984. The evolutionary
sig-
nificance of
genetic
diversity:
Ecological,
demographic
and life
his-
tory
coreletes. In
G. S. Mani [ed.],
Evolutionary dynamics
of genetic
diversity,
13-213. Springer-Verlag,
Berlin.
SCANDALIOS, J.
G. 1969. Genetic
control of
multiple
molecular
forms
of
enzymes
in
plants: a review.
Biochemical
Genetics
3: 37-79.
SCHNABEL, A., AND J.
L. HAMRICK. 1990. Comparative
analysis of
population
genetic
structure
in
Quercus
macrocarpa
and Q.
gambelii
(Fagaceae). Systematic
Botany 15: 240-251.
SCHUSTER, W. S., D. L. ALLES, AND J.
B. MITTON. 1989. Gene flow
in
limber
pine:
evidence from
pollination
phenology
and genetic
dif-
ferentiation
along
an
elevational transect.
American
Journal
of
Bot-
any
76: 1395-1403.
SCHWARZMANN,
J. F., AND H. D. GERHOLD. 1991. Genetic
structure
and mating system of Northern Red Oak (Quercus rubra
L.) in
Pennsylvania.
Forest
Science 37: 1376-1389.
SLATKIN, M. 1981. Estimating
levels
of
gene
flow
in natural
popula-
tions.
Genetics 99: 323-335.
1985. Rare alleles as indicators of
gene flow.
Evolution 39:
53-65.
SOLTIS, D. E., C. H. HAUFLER, D. C. DARRow, AND J.
GASTONY. 1983.
Starch
gel
electrophoresis
of ferns: a
compilation
of
grinding
buffers,
gel and electrode
buffers,
and staining
schedules. American
Fern
Journal 73: 9-27.
SURLES, S. E.,
J.
L. HAMRICK, AND B. C. BONGARTEN. 1989. Allozyme
variation in black locust Robinia
pseudoacacia. Canadian Journal
of
Forestry
Research 19: 471-479.
SwOFFoRD, D. L., AND
R. B. SELANDER. 1989. BIOSYS-1: a computer
program
for
the
analysis
of
allelic variation
in
population
genetics
and biochemical
systematics,
release 1.7. Illinois
Natural
History
Survey,
Champaign,
IL.
VALLEJOS,
C. E. 1983. Enzyme
activity
staining.
In S. D. Tanksley
and
November 1993] HOKANSON ET AL. -ISOZYME VARIATION IN OAKS OF THE APOSTLE ISLANDS 1357
T. J. Orton
[eds.],
Isozymes
in Plant Genetics
and Breeding,
part
A, 469-516. Elsevier
Science,
Amsterdam.
WEIR, B. S. 1989. Sampling properties
of gene
diversity.
In A. H. D.
Brown,
M. T. Clegg,
H. L. Kahler,
and B. S. Weir [eds.], Plant
population
genetics, breeding
and genetic resources,
23-42. Sinauer,
Sunderland,
MA.
WHITTEMORE, A. T., AND B. A. SCHAAL. 1991. Interspecific
gene
flow
in sympatric
oaks.
Proceedings
of the
NationalAcademy
ofSciences,
USA 88: 2540-2544.
WRIGHT, S. 1931. Evolution
in mendelian
populations.
Genetics 16:
97-159.
1951. The genetical
structure
of
populations.
Annals of Eu-
genics
15:
323-354.
. 1965. The
interpretation
of
population
structure
by
F-statistics
with
special
regard to
systems of
mating.
Evolution
19: 395-420.
